Instructions to use 0xrushi/chesslfm-experiments with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use 0xrushi/chesslfm-experiments with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="0xrushi/chesslfm-experiments")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("0xrushi/chesslfm-experiments", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use 0xrushi/chesslfm-experiments with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "0xrushi/chesslfm-experiments"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "0xrushi/chesslfm-experiments",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/0xrushi/chesslfm-experiments

SGLang

How to use 0xrushi/chesslfm-experiments with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "0xrushi/chesslfm-experiments" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "0xrushi/chesslfm-experiments",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "0xrushi/chesslfm-experiments" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "0xrushi/chesslfm-experiments",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use 0xrushi/chesslfm-experiments with Docker Model Runner:
```
docker model run hf.co/0xrushi/chesslfm-experiments
```

0xrushi commited on 5 days ago

Commit

d310ecc

verified ·

1 Parent(s): c845cdf

Clarify LiquidAI to ChessLFM model lineage

Browse files

Files changed (1) hide show

self-play-refresh/README.md +4 -0

self-play-refresh/README.md CHANGED Viewed

@@ -13,3 +13,7 @@ Fine-tuned after one self-play data-refresh round, mixing student-reached positi
 Released reference scored 15/32 in this block. This result is not directly comparable with the later-block evaluations.
 This checkpoint uses the ChessLFM chess-specific tokenizer. For legal move masking, board prompt construction, and search, see the project implementation at https://github.com/0xrushi/ChessLFM.

 Released reference scored 15/32 in this block. This result is not directly comparable with the later-block evaluations.
 This checkpoint uses the ChessLFM chess-specific tokenizer. For legal move masking, board prompt construction, and search, see the project implementation at https://github.com/0xrushi/ChessLFM.
+## Model lineage
+Direct starting checkpoint: [mlabonne/LFM2.5-230M-Chess](https://huggingface.co/mlabonne/LFM2.5-230M-Chess). That upstream checkpoint was itself built from [LiquidAI/LFM2.5-230M-Base](https://huggingface.co/LiquidAI/LFM2.5-230M-Base). This experiment therefore descends from LiquidAI's pretrained base through ChessLFM; it did not start directly from the LiquidAI base weights or from random initialization.